The wait continues for report on Lydiard St rail gate smash

 
Topic moved from News by bevans on 28 Jul 2020 10:12
  historian Deputy Commissioner

Brings me back to the initial point. The gates could have been opened, there was time (the other booms did work). It didn't happen due to a multitude of events.
justarider

If there had been booms at Lydiard St they would have had time for one, perhaps two, dings before the train was over the crossing. The booms wouldn't have even started to descend.

Signalling practice in Victoria is such that where there is a platform on the approach side a level crossing special interlocking is provided between the level crossing protection equipment, the protecting signal, and the approach track circuit. Where a train is to stop at the platform, the protecting signal is held at stop and this cuts out the approach track circuit. The effect of this is that a train can stop at the platform to do station work without activating the boom barriers/flashing lights. When the train should move off, the boom barriers commence to operate. When they are proved down, the protecting signal clears, and the train can depart. (I've glossed over some of the not relevant details.)

In this case, the signaller would have held the home at the end of the platform at stop as the approaching train was scheduled to stop at Ballarat. The boom barriers would not have commenced to operate until the first axle of the runaway train had passed the home signal - from memory this is about 10 metres from the crossing. Track relays typically take about a second to drop so the train would have been on the level crossing before the signalling system even registered it was there (100 km/h = 28 m/s). At just under 100 km/h, the train would have cleared the level crossing track circuit before the protection equipment had got up to the point of allowing the booms to drop.

The only advantage of booms in this case would have been avoiding the timber fragments being sprayed around. This is not a minor advantage, of course, the timber fragments would have easily killed pedestrians if any had been around.

Sponsored advertisement

  justarider Chief Commissioner

Location: Free at last, free at last
Brings me back to the initial point. The gates could have been opened, there was time (the other booms did work). It didn't happen due to a multitude of events.
"justarider"

If there had been booms at Lydiard St they would have had time for one, perhaps two, dings before the train was over the crossing. The booms wouldn't have even started to descend.

Signalling practice in Victoria is such that where there is a platform on the approach side a level crossing special interlocking is provided between the level crossing protection equipment, the protecting signal, and the approach track circuit. Where a train is to stop at the platform, the protecting signal is held at stop and this cuts out the approach track circuit. The effect of this is that a train can stop at the platform to do station work without activating the boom barriers/flashing lights. When the train should move off, the boom barriers commence to operate. When they are proved down, the protecting signal clears, and the train can depart. (I've glossed over some of the not relevant details.)

In this case, the signaller would have held the home at the end of the platform at stop as the approaching train was scheduled to stop at Ballarat. The boom barriers would not have commenced to operate until the first axle of the runaway train had passed the home signal - from memory this is about 10 metres from the crossing. Track relays typically take about a second to drop so the train would have been on the level crossing before the signalling system even registered it was there (100 km/h = 28 m/s). At just under 100 km/h, the train would have cleared the level crossing track circuit before the protection equipment had got up to the point of allowing the booms to drop.

The only advantage of booms in this case would have been avoiding the timber fragments being sprayed around. This is not a minor advantage, of course, the timber fragments would have easily killed pedestrians if any had been around.
"historian"

Thanks @historian for a great description of what happens in a normal event.

Unfortunately this was a very ab-normal occurrence. The reaction to it was revealed wanting.
A stop signal in front of a train that clearly cannot stop, is not going to help.

In the realm of what ifs, the signaller could have changed the approach & home to proceed and allowed booms to activate accordingly.
Of course it was gates, remote controlled, so reaction time was more critical if even possible.

cheers
John
  Galron Chief Commissioner

Location: Werribee, Vic
Are/where the signals at Ballarat fitted with TWPS? Going forward, could you use that and perhaps have the signals detect the train passing some point to fast and take some pre-determine action?

Also, I thought placing the train in emergency put a flag up on the train controllers panel? Would have thought that would have had train control on the radio? And ICE radios I thought also have/had an emergency button on them to do the same?
  kitchgp Chief Commissioner

It was 10 seconds from the time the Emergency brake was applied until the VLocity passed the Departure signal at Stop. It passed through the crossing 2 seconds later.
  historian Deputy Commissioner

In the realm of what ifs, the signaller could have changed the approach & home to proceed and allowed booms to activate accordingly.
Of course it was gates, remote controlled, so reaction time was more critical if even possible.
justarider

Bear in mind that the driver has to realise that they have a problem, contact the signaller, communicate the problem, for the signaller to understand, for the signaller to react and operate the controls, for the telemetry system to communicate the instructions to the interlocking at Ballarat and the interlocking to respond. That's quite a delay when the train is approaching at 100 km/h.

The answer depends on whether the hypothetical boom barriers were manually or automatically controlled (the gates were manually controlled). Manual control from a dedicated switch was only provided where there was a lot of shunting and allowed the signaller to make successive moves without waiting for the booms to cycle each time. If the booms had been provided when Ballarat was resignalled, and movements were loco hauled trains, manual control would have probably been provided. Now, probably not.

If manual control was provided, the signaller would only need to operate the control. The booms would commence operating.

If manual control was not provided, operation would be by occupancy of the control track circuit approaching the level crossing & suppressed by the home at stop. The length of the track circuit would be set to give about 25 seconds operation of the booms at maximum line speed before the level crossing is occupied. Given the line speed at Lydiard St is 40 km/h, but this train approached at 100 km/h, the maximum warning time would be about 10 seconds.

From the driver's perspective, if the signaller had managed to call the signal before the train entered the control track circuit, the signal would clear immediately and the booms would start when the train entered the control track circuit. In this case 10 seconds before reaching the level crossing. If the train was already on the control track circuit when the signaller called the signal, the booms would start operating immediately, but the signal would remain at stop. It would not clear until the booms were proved horizontal - since the train would pass the signal before this happened, the driver would pass a red signal. By definition this would give less than 10 seconds operating time.

The signal could only be called if the points at the Ararat end were lying for No 1 Road. I assume they were as there is no mention of trailing through the points. If they weren't, the signaller couldn't call the signal (and cause the booms to operate) until the points had been reversed. Add many seconds to the time.
  historian Deputy Commissioner

Are/where the signals at Ballarat fitted with TWPS? Going forward, could you use that and perhaps have the signals detect the train passing some point to fast and take some pre-determine action?
Galron

No. TPWS is a train bourne system. It allows a train to detect that it has passed a point travelling too fast (and apply the brake), but it does not communicate this information to the signalling system.

The only signal at danger the train passed was the one at the end of the platform, just before entering the level crossing. By this stage the brakes had already been in emergency for 10 seconds, so TPWS is moot.
  kitchgp Chief Commissioner

The “remote controlledness” of the gates just means they are no longer operated by a wheel in the signal box, but are motor driven and operated from a panel, now located in the station and/or at Centrol.  If boom gates were installed, they probably would be controlled in the same manner.

Apart from demolishing the gates, some other equally, or perhaps, more important aspects of the incident were:
  • the VLocity didn’t derail when diverging at 100 km/h over 40 km/h points, approaching the station,
  • it didn’t injure platform passengers as it passed through at an unexpected 100 km/h,
  • there weren’t any trucks, buses, cars or pedestrians on the crossing (booms or gates, makes no difference), and
  • there wasn’t a train, such as a VLocity from Maryborough or Wendouree, coming the other way. It stopped ~500 m past the Departure signal, well into single-track territory.
  •   hbedriver Assistant Commissioner

    The lack of a serious collision or worse was a blessing.

    Emergency brakes applied at about 100km/h, train then passed through tight curves (as in, high friction against the train), then ran uphill and stopped after around 800m. Given that we stop V’locity trains in service brakes from that speed  in half that distance, pretty hard to blame the driver.

    Driver very experienced, former instructor and had thousands of runs there on that train type. Tens of thousands of trips down that hill counting steam and diesel locos. I usually apply brakes about where he initially did; beats me why it didn’t work that time. Can only assume brake computer or wheel slide system failure.

    Reported that no loud or obvious wheel flats on the train, so skidding seems unlikely, perhaps the equipment was too busy trying to protect the wheels from flats that stopping became a secondary consideration?
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    '.... perhaps the equipment was too busy trying to protect the wheels from flats that stopping became a secondary consideration?'
    That has happened before.
    A Budd car piled into the dead end at Port Pirie many years ago when the wheel slide system 'thought .............'.
      Valvegear Oliver Bullied, CME

    Location: Richmond Vic
    "Brake computer" ! ! !  A combination of the driver's skill and the Westinghouse Automatic air Brake worked well for decades. A brake computer is like a steering column gearshift - a complicated way of doing something simple. Contrary to what some people believe, a computer is not the ultimate means of control - just ask Boeing about the 737-MAX for example.
      historian Deputy Commissioner

    The lack of a serious collision or worse was a blessing.

    Emergency brakes applied at about 100km/h, train then passed through tight curves (as in, high friction against the train), then ran uphill and stopped after around 800m. Given that we stop V’locity trains in service brakes from that speed  in half that distance, pretty hard to blame the driver.

    Driver very experienced, former instructor and had thousands of runs there on that train type. Tens of thousands of trips down that hill counting steam and diesel locos. I usually apply brakes about where he initially did; beats me why it didn’t work that time. Can only assume brake computer or wheel slide system failure.

    Reported that no loud or obvious wheel flats on the train, so skidding seems unlikely, perhaps the equipment was too busy trying to protect the wheels from flats that stopping became a secondary consideration?
    hbedriver

    Indeed.

    It is notable that in the first application of the brake on Warrenheip bank, the driver was able to reduce the speed from 160 km/h to 141 km/h in 20 seconds. About 1km/h/s on a 1 in 52 falling grade, with wheel slide activated and with sand.

    On the flatter section at the foot of the grade in the second application the driver was only able to reduce speed from 160 km/h to 117 km/h in 68 seconds. About 0.63 km/h/s on a much flatter grade, again with wheel slide activated and with sand.

    Of course the report doesn't give us how hard the driver was braking in these two instances.

    In the next 14 seconds the speed fell further to 99 km/h - the decelaration rate increased to nearly 1.3 km/h/s. In 14 seconds at an average speed of 108 km/h the train would have travelled 420 metres. Given that this distance was measured from about Humffray St, it looks like this stretch would have included the crossover to No 1 Road, which suggests that some of this deceleration was due to the energy being absorbed in the traversal of the crossover.

    It's also notable that during the second brake application the wheel slide indicator remained on for essentially the entire stop, while the sanding cut out after 71 seconds. If I've counted the time reported correctly, this would be four seconds before the driver put it into emergency, and right about the time the train was probably going through the crossover.

    Twelve seconds later the speed had fallen 6 km/h - a decelaration of 0.5 km/h/s on level track with the brake in emergency, wheel slide on, but apparently no sand. At roughly the end of this period the train destroyed the gates and went through the turnout, both of which would have absorbed energy. So this decelation may have included part of this.

    The train came to a stand 39 seconds later - a decelaration of 2.4 km/h/s, with the brakes in emergency, wheel slide still on, no sand, but on a gently rising grade, partially on a sharpish curve, and probably including the travesal of the turnout, and perhaps the destruction of the gates.

    When looking at these deceleration rates, kinetic energy is proportional to the square of the speed while the ability of the brake to absorb energy is usually constant and independent of speed. In practice this means that the deceleration rate is slower at higher speeds and faster at lower.

    Finally, around the time of the Siemens broo-ha-ha, I found an RAIB report into an incident in the UK (I can't, of course, find it now). In this report a multi-car DMU had run away for several miles on an icy track due to the wheel slip/slide system releasing the brakes. While these systems are often equated to ABS on cars, the report strongly implied that purpose of wheel slip/slide systems was not primarily to reduce the distance to stop by preventing sliding, but rather to prevent wheel flats and consequent maintenance costs. With the DMU in the RAIB report, the wheel slip/slide system had essentially disabled the brake for several miles.

    I would assume the final report would look at the wheel slip/slide system, the sanding system, the interaction of both with braking system modes, and the drivers' knowledge of all of this.
      historian Deputy Commissioner

    Finally, around the time of the Siemens broo-ha-ha, I found an RAIB report into an incident in the UK (I can't, of course, find it now). In this report a multi-car DMU had run away for several miles on an icy track due to the wheel slip/slide system releasing the brakes. While these systems are often equated to ABS on cars, the report strongly implied that purpose of wheel slip/slide systems was not primarily to reduce the distance to stop by preventing sliding, but rather to prevent wheel flats and consequent maintenance costs. With the DMU in the RAIB report, the wheel slip/slide system had essentially disabled the brake for several miles.
    historian

    And, of course, I found the RAIB report as soon as I posted...

    https://assets.publishing.service.gov.uk/media/547c8fe7e5274a428d000159/R182011_111117_Stonegate.pdf

    I must have mixed the report up with other reading at the time, as it doesn't talk about wheel slip/slide systems (though the train was fitted with it), but does discuss extensively low adhesion and sanding problems.

    The incident, in fact, sounds very similar to the Ballarat accident.
      Valvegear Oliver Bullied, CME

    Location: Richmond Vic
    Thanks for all of your posts, historian. They are invariably extremely detailed and informative, and I'm sure a lot of others would agree.
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    The underlying cause/causes of this incident must be known already because if they are not Vlocitys must be regarded as unsafe in some circumstances at least. Especially where the safety margin is the thickness of a signal post.
    Are they Vline's version of the 737MAX? I don't know but the ATSB Report needs to be expedited to allay passenger concerns.
    (Re-submit in 2024.)
      The Vinelander Minister for Railways

    Location: Ballan, Victoria on the Ballarat RFR Line
    The underlying cause/causes of this incident must be known already because if they are not Vlocitys must be regarded as unsafe in some circumstances at least. Especially where the safety margin is the thickness of a signal post.
    Are they Vline's version of the 737MAX? I don't know but the ATSB Report needs to be expedited to allay passenger concerns.
    (Re-submit in 2024.)
    YM-Mundrabilla

    The 6km in length speed restriction of 80Km/h down the Warrenheip bank to Ballarat East, well and truly allays the passengers concerns.

    Mike.
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    The underlying cause/causes of this incident must be known already because if they are not Vlocitys must be regarded as unsafe in some circumstances at least. Especially where the safety margin is the thickness of a signal post.
    Are they Vline's version of the 737MAX? I don't know but the ATSB Report needs to be expedited to allay passenger concerns.
    (Re-submit in 2024.)

    The 6km in length speed restriction of 80Km/h down the Warrenheip bank to Ballarat East, well and truly allays the passengers concerns.

    Mike.
    The Vinelander
    Not much help in this case. Did not the speed actually increase/barely decrease on this section regardless of the speed over the top of Warrenheip.
      kuldalai Chief Commissioner

    Somehow this all sounds very similar to an incident when I was a teenager on Warrenheip Bank, with a single T class (which were notoriusly poor braking locos to start with).  So a single T class heading Down Warrenheip bank with a long rake of 4 wheel GY wagons lost braking control. Smashed through the then interlocked gates at Ballarat East (Hummfray Street ?) and the empty GY's  just piled up in a Mt Fuji like pyramid of derailed GY wagons.

    Similar incident different location, different cause at Werribee likewise many years ago.  Again GY wagons, broken W guard (wheel support) passing through Werribee creating a mountain of derailed GY's again this time right under the old footbridge linking the platforms.

    Good old VR days of rakes of 75 x 4 wheel GY wagons used on grain trains, with all the grain being hand shovelled out at GEB North Geelong.
      historian Deputy Commissioner

    The underlying cause/causes of this incident must be known already because if they are not Vlocitys must be regarded as unsafe in some circumstances at least. Especially where the safety margin is the thickness of a signal post.
    Are they Vline's version of the 737MAX? I don't know but the ATSB Report needs to be expedited to allay passenger concerns.
    (Re-submit in 2024.)

    The 6km in length speed restriction of 80Km/h down the Warrenheip bank to Ballarat East, well and truly allays the passengers concerns.
    Not much help in this case. Did not the speed actually increase/barely decrease on this section regardless of the speed over the top of Warrenheip.
    YM-Mundrabilla

    The speed restriction, apparently, is for SPAD mitigation Smile

    Halving the speed down Warrenheip bank reduces the kinetic energy required to be handled by the brakes to a quarter of that during the descent being discussed. The potential energy added during the descent remains the same, of course, but the brakes have twice as long to dissipate it.

    All that means that the demands on the brakes are far, far, less. Less demand on the brakes means the demands on the wheel/rail interface are reduced in the same proportion.
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    Somehow this all sounds very similar to an incident when I was a teenager on Warrenheip Bank, with a single T class (which were notoriusly poor braking locos to start with).  So a single T class heading Down Warrenheip bank with a long rake of 4 wheel GY wagons lost braking control. Smashed through the then interlocked gates at Ballarat East (Hummfray Street ?) and the empty GY's  just piled up in a Mt Fuji like pyramid of derailed GY wagons.

    Similar incident different location, different cause at Werribee likewise many years ago.  Again GY wagons, broken W guard (wheel support) passing through Werribee creating a mountain of derailed GY's again this time right under the old footbridge linking the platforms.

    Good old VR days of rakes of 75 x 4 wheel GY wagons used on grain trains, with all the grain being hand shovelled out at GEB North Geelong.
    kuldalai
    Similar 'lost the air' on up goods trains (usually T class) coming down Glenroy were not all that infrequent either. OK if you had the distant for Pascoe Vale in which case just hold on - otherwise hold on to the whistle cord.
      Lockspike Chief Commissioner

    Somehow this all sounds very similar to an incident when I was a teenager on Warrenheip Bank, with a single T class (which were notoriusly poor braking locos to start with).  So a single T class heading Down Warrenheip bank with a long rake of 4 wheel GY wagons lost braking control. Smashed through the then interlocked gates at Ballarat East (Hummfray Street ?) and the empty GY's  just piled up in a Mt Fuji like pyramid of derailed GY wagons.

    Similar incident different location, different cause at Werribee likewise many years ago.  Again GY wagons, broken W guard (wheel support) passing through Werribee creating a mountain of derailed GY's again this time right under the old footbridge linking the platforms.

    Good old VR days of rakes of 75 x 4 wheel GY wagons used on grain trains, with all the grain being hand shovelled out at GEB North Geelong.
    kuldalai
    Was that the incident where an air tap wasn't opened again after shunting at a wayside station, and the only operating brakes were on the loco? IIRC, the signalman diverted the train into a dead end road in the yard?
      Galron Chief Commissioner

    Location: Werribee, Vic
    I may have missed it, but has rail head contamination been ruled out? Not unheard of for people to think its funny and grease up a rail head? although over the distances here, it might be a bit much to blame that entirely unless it was some natural contamination?
      historian Deputy Commissioner

    I may have missed it, but has rail head contamination been ruled out? Not unheard of for people to think its funny and grease up a rail head? although over the distances here, it might be a bit much to blame that entirely unless it was some natural contamination?
    Galron

    As the ATSB report puts it...

    The investigation is continuing and will include the review and examination of... evidence in relation to rail condition and wheel/rail interface

    In this context I would note the season that this incident occurred in (albeit just).

    My view? Some form of rail head contamination is almost certain to be a casual factor. But not deliberate. And it won't be the whole story by any chalk.
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    Rail accidents/incidents are almost always a combination of often minor, in isolation, issues such as:

    • No sand/ineffective sanders
    • Wheel profile
    • Rail head condition
    • Rolling stock maintenance
    • Train handling
    • Wheel slip/slide computer 'confusion' and/or fault

    The foregoing are but a few.
      YM-Mundrabilla Minister for Railways

    Location: Mundrabilla but I'd rather be in Narvik
    Finally, around the time of the Siemens broo-ha-ha, I found an RAIB report into an incident in the UK (I can't, of course, find it now). In this report a multi-car DMU had run away for several miles on an icy track due to the wheel slip/slide system releasing the brakes. While these systems are often equated to ABS on cars, the report strongly implied that purpose of wheel slip/slide systems was not primarily to reduce the distance to stop by preventing sliding, but rather to prevent wheel flats and consequent maintenance costs. With the DMU in the RAIB report, the wheel slip/slide system had essentially disabled the brake for several miles.

    And, of course, I found the RAIB report as soon as I posted...

    https://assets.publishing.service.gov.uk/media/547c8fe7e5274a428d000159/R182011_111117_Stonegate.pdf

    I must have mixed the report up with other reading at the time, as it doesn't talk about wheel slip/slide systems (though the train was fitted with it), but does discuss extensively low adhesion and sanding problems.

    The incident, in fact, sounds very similar to the Ballarat accident.
    historian
    Too many computers, too many players between the lot of which no one and no system told the driver that he was low/had no sand. This would almost certainly have led him to modify his driving technique. The computer somewhere and a cast of engineers and others knew but not the bloke at the front. I assume that no one does any sort of 'in service' visual these days on anything.
    I got sick of reading after about page 25 but saw nothing mentioned about wheel profile up to that point.
      route14 Chief Commissioner

    When I questioned about the position of sanders on linkedin, someone told me to have a second look, but they are illustrated behind the leading bogie whichever end it's driven from.  Is there a different illustration?

    Sponsored advertisement

    Display from: